Estimation of adult census size from close-kin dyads in the malaria mosquito Anopheles gambiae

Accurate estimates of adult mosquito abundance are central to the design and evaluation of vector control strategies, yet they are difficult to obtain from natural populations. Conventional mark-recapture methods for estimating adult mosquito census size pose logistical and other challenges. A recently developed close-kin mark-recapture (CKMR) approach is a promising alternative. However, application of CKMR has been largely confined to long-lived vertebrates. Validation on empirical data from short-lived, highly fecund insects is lacking. Here, we apply CKMR to a natural population of Anopheles gambiae, the primary African malaria mosquito, sampled from a small island in Lake Victoria, Uganda. Using a high-diversity amplicon panel of genome-wide markers, we genotyped 714 adult mosquitoes collected over a 20-day period. We classified pairs to close-kin categories by implementing probabilistic latent-kinship estimation rather than using a deterministic threshold-based framework. We observe numerous full-sibling pairs but no parent-offspring pairs, a pattern indicative of extreme variance in reproductive success owing to frequent failure of mosquito egg clutches to produce adults. To adapt the CKMR framework to mosquito life history, we explicitly modeled the mosquito life cycle including the possibility of clutch failure. By incorporating probabilities of both parent-offspring and full-sibling relationships, we estimate an adult female census size of 26,887 (95% credible interval: 6,979 - 146,011), and a clutch failure probability of 97.6% (CrI: 91.3 - 99.6%). We use these estimates to calculate the predicted variance in reproductive success and effective population size. Through individual-based simulations we confirm our estimates and the necessity of modeling reproductive variance to explain observed kinship patterns. Our results demonstrate that CKMR can be applied successfully to mosquitoes, provided that appropriate adjustments are made to account for their natural history.精准估算成蚊种群数量，是媒介防控策略设计与评估的核心环节，但从自然种群中获取此类数据却颇具挑战。传统用于估算成蚊种群总数的标记重捕法，存在后勤及其他方面的诸多难题。近年来新兴的近缘个体标记重捕法（close-kin mark-recapture, CKMR），便是颇具潜力的替代方案。然而目前CKMR的应用大多局限于长寿命脊椎动物，针对短寿命、高繁殖力昆虫的实证数据验证仍存在空白。本研究将CKMR方法应用于采自乌干达维多利亚湖一座小岛的自然冈比亚按蚊（Anopheles gambiae）种群——该蚊种是非洲主要的疟疾传播媒介。本研究采用覆盖全基因组的高多样性扩增子标记面板，对20天内采集的714只成蚊进行基因分型。我们未采用基于确定性阈值的分类框架，而是通过概率性潜在亲缘关系估算，将蚊样配对划分为不同近缘类群。研究观察到大量全同胞配对，但未发现亲子配对，这一模式表明，由于蚊卵簇常无法发育为成蚊，其繁殖成功率存在极强的变异度。为使CKMR框架适配蚊类生活史特征，本研究对蚊类生命周期进行了显式建模，其中纳入了卵簇发育失败的可能性。通过整合亲子关系与全同胞关系的概率，我们估算得到成雌蚊种群总数为26887（95%置信区间：6979~146011），卵簇发育失败概率为97.6%（95%置信区间：91.3%~99.6%）。基于上述估算结果，我们计算了繁殖成功率的预测变异度与有效种群大小。通过个体水平模拟，我们验证了上述估算结果，并证实了为解释观测到的亲缘关系模式，建模繁殖变异度的必要性。本研究结果表明，只要针对蚊类的自然生活史做出恰当调整，CKMR方法便可成功应用于蚊类种群研究。